Survival – Performing Log-Rank Test After Optimal Pair Matching

Question

I have clinical data and used optimal matching in the MatchIt package to match cases to controls on several variables. Matching was done in a 1:1 ratio, and balance was achieved.

I then did a Kaplan-Meier analysis with a log rank test in the R survival package. A reviewer has said I need "the log rank test should be stratified by pairs."

I'm unclear why this is necessary – the only weights I'm seeing from the matching procedure are 0's and 1's.

Is there a way to "stratify by pairs" in R?

Any guidance to references is appreciated!

Answer 1

Optimal matching returns pair membership in the subclass column of the match.data() output. This can be added to the log-rank test as a stratum variable using the following model formula:

Surv(time, status) ~ treat + strata(subclass)

This is recommended to account for the paired nature of the matched sample and has nothing to do with the weights.